Replication Project Part 03: Bias Mitigation
Class Coverage
- Mitigating fairness issues in AI through pre-, in-, and post-processing
- Re-learning and re-deploying models.
Pre-Class Readings
Please read the following:
- Model Cards for Model Reporting (Mitchell et al.)
- Fairness Through Awareness (Dwork et al.)
- Demo: Explaining model behavior using LIME
Participation Questions
Submit to Gradescope by 11AM PT on , Thursday November 16th
- SUBMIT DEFAULT ACTIVITY PARTICIPATION QUESTIONS
- In what situations do you think “fairness through awareness” is most appropriate? Are there any situations where “fairness through unawareness” is necessary instead?
- Describe a situation in which a pre-, in-, or post-processing technique might be appropriate for mitigating model bias, and explain what technique you would use.
This Week’s Slides
Assigned for Week 08
- Complete Replication Project Part 03a
- Complete Replication Project Part 03b - Write up Instructions
- Complete next week’s readings
- If you signed up to present on Algorithmic Fairness and Vertical Equity (Black et al.) come prepared to present next week and submit your presentation on Gradescope by 10AM PT on Friday, November 18th
- Submit your answers to next week’s participation questions to Gradescope by 10AM PT on Thursday, November 17th
Office Hours for Replication Project
- Tuesdays from 3:30-4:30pm PT
- Reach out on Discord for questions. @ Parker Addison or Emily Ramond.